RFC 227

RFC 227 💖 DATA TRANSFER RATES (RAND/UCLA)

okay bestie, the header facts first! 🎀 This one is written by J. Heafner and E. Harslem (NIC 7631), dated September 17, 1971. It updates RFC 113 ✨. The original doesn't list a category or say what it obsoletes, so I'm not making that up 🙈

Okay so this RFC is basically a lowkey data-rate diary from RAND and UCLA 🧚 It's about how fast data moves over the network, and why it's sometimes slower than it should be. The rates in the attached memo are what's typical for how we use RJS at UCLA CCN 💅

Earlier timing tests with UCSB (similar, but more detailed) showed that most of the time got wasted because of:

1. channel contention with our disk drive access 💀
2. our NCP runs at a higher priority than batch jobs, but lower than text editing and interactive graphics 🙄
3. OS interrupt handling is very slow on both ends 😭
4. spooling time of the remote system 🫠

The Memo 📝

Here's the memo, no cap. It's giving office drama energy 🍓

TO:      John Heafner
FROM:    Bob Hoffman
COPIES:  Bob Mobley, Herb Shukiar

Bob noted these transmission rates over the network between Rand and UCLA. They were all taken at night, when not much else was going on with our 65 🌙

Send to UCLA 🚀

   # Cards    Blocksize (bytes)  Time (secs)   Rate (bits/secs)
      642        80                   50             8218
      375        80                   30             8000
      509       800                   20            16288

Receive from UCLA 📥

For every figure in this part, the receiving file has a blocksize of 1330 bytes, and each line is assumed to hold 100 bytes. That second assumption is pretty accurate, since most of the lines came from PL/I, where 100 is a really good number 🎀 So for each rate, the number of bytes is the # Lines * 100. Math ate, bestie 💅

   # Lines          Time (secs)      Rate (bits/secs)
      4900              200             19600
       872               47             14843
      3900              185             16865

The Takeaway 🔥

Okay, the main vibe from the send numbers: blocking makes about a 2:1 difference 👑 Memory also remembers a 2 or 3 to 1 advantage for blocking on receive, back when we were getting unblocked files from UCSB 🧠✨

REH:gb 🫶

Machine-Readable Note 🤖

[This RFC was put into machine readable form for entry] [into the online RFC archives by Kelly Tardif, Viag�nie 10/99] 🌈